Jun 2022 – Apr 2024 · 1 yr 10 mos · Pune, Maharashtra, India

• Full-Stack Ecosystem Ownership: Exercised end-to-end technical ownership over a mission-critical suite including CrmService, CartService, NotificationService, TicketService, Crm-UI, and Offer-UI.
• B2B Logic & Event-Driven Systems: Designed and implemented an asynchronous Partner Club Assignment engine (5-tier logic) based on purchase and payment behavior.
• Critical Infrastructure Stability: Identified and resolved a legacy database "hang-up" issue affecting the Django-based support portal.
• Latency Reduction: Optimized UI performance during the "Credit Cash" rollout by building efficient Aggregator Layers for Catalog and Cart service endpoints, resolving significant lag caused by dual-cart management.
• CI/CD Transformation: Overhauled the deployment pipeline for Crm-UI and Offer-Management-UI by implementing CI/CD, reducing deployment cycles from 20 minutes (manual build/restart) to under 1 minute.
• Infrastructure Modernization: Led the migration of frontend build systems to Node.js v16 and improved system observability by performing comprehensive error-log cleanups across core backend services.
• Multi-Channel Communication Strategy: Scaled the notification infrastructure by integrating third-party WhatsApp and SMS gateways, enabling a strategic architectural split between B2C and B2B communication channels.
• High-Velocity Execution: Demonstrated elite delivery speed by leading the "Surkhiyan" project from prototype to production in 5 days and delivering "Weather Updates" within 48 hours.
• Recognized Excellence: Consistently rated as a "7-Star Performer" and "Impact Creator," receiving 3 consecutive quarterly performance awards for high-impact contributions and technical mentorship.
• Stakeholder Alignment: Actively drove architecture discussions and code reviews, ensuring high standards for code quality and 100% KPI achievement for project velocity and risk mitigation.

Oct 2021 – May 2022 · 7 mos · Pune, Maharashtra, India

• Product Architecture & Decision Engines
• Rule-Based Recommendation Engine: Architected a dynamic, state-driven UI engine using Node.js and React/Redux that managed complex insurance recommendation workflows.
• Anonymous-to-Authenticated Data Mapping: Designed a session-merging strategy that seamlessly mapped "Guest" session data and recommendation history to user profiles upon post-recommendation login, ensuring zero data loss during the conversion funnel.
• Intelligent Provider Filtering: Integrated a machine learning feedback loop that filtered insurance providers based on historical rejection patterns for specific user archetypes. This optimization increased successful policy issuance by 20%.
• State Persistence & Session Management: Implemented a robust server-side and client-side (Redux) state management strategy to handle high-frequency session updates and complex insurance plan comparisons across disparate providers.
• Operational Impact & Growth
• User Behavior Analytics: Strategized and implemented an end-to-end user activity tracking system to identify high-friction drop-off points in the funnel, leading to data-driven UI/UX refinements.
• Referral System Design: Engineered a viral growth loop by building a referral functionality from scratch, enabling tracked user invitations and reward attribution within the recommendation platform.
• Hybrid Fulfillment Logic: Developed a "manual fallback" workflow where low-confidence ML recommendations automatically triggered a "Buddy Assignment" system, ensuring 100% lead coverage by connecting high-risk users with insurance agents for manual follow-up.

Jul 2018 – Sep 2021 · 3 yrs 2 mos · Pune Area, India

• Distributed Systems & Scalability
• Architectural Transformation: Re-engineered a monolithic, cron-based data ingestion pipeline into a distributed microservices architecture (Node.js, Python, Celery), reducing processing time by 90% (from 30 mins for 100 posts to ~5 mins for 400+ posts).
• High-Concurrency Task Orchestration: Designed a multi-stage processing pipeline using Celery and RabbitMQ, implemented a "RootService" to chunk massive social media datasets into parallel workloads.
• Auto-Scaling Infrastructure: Developed a dynamic scaling mechanism for "AiService" workers using AWS EC2 custom images to spin up/down compute resources on-demand based on real-time Celery task queue depth, optimizing throughput.
• Granular State Management: Engineered a robust tracking system to monitor progress at the individual post, fetch, and profile levels, ensuring data consistency and enabling real-time status updates for the UI.
• Data Engineering & Product Impact
• ML-Driven Social Intelligence: Built the core logic for an analytics platform that ingested Facebook/Instagram Graph API data to provide predictive "when and what to post" reports based on historical engagement patterns.
• Full-Stack Ownership: Acted as the sole Python Developer for core services while contributing significantly to the Node.js API and React/AngularJS frontend to bridge the gap between heavy background processing and data visualization.
• System Optimization: Improved platform reliability by replacing manual database dumps with event-driven task queues, facilitating a seamless flow of data from ingestion to final chart rendering.
• MVC Development: Led the technical execution of a Minimum Viable Product (MVP), collaborating directly with stakeholders to translate business requirements into a scalable technical specification.